Structured digital predistorter model derivation based on iterative learning control

A novel approach to derive model structures for digital predistorters based on iterative learning control (ILC) is presented. ILC is a technique used to identify the optimal PA input signal/predistorted signal that drives a PA to a desired output response. The ILC concept is used to derive an analytical expression of the predistorted signal and to identify basis functions for predistorter models. The proposed approach is used to derive a predistorter model structure from the memory polynomial model. Experimental results show that the predistorter model derived with the proposed approach can obtain better linearity performance than conventional models used in digital predistortion.